Guess how much AR widowmaker got

[crattey]

SL mobs don't have evades, just AAD. Not sure about RK mobs though.

[PingBong]

ok so i decided to recalculate the AR of as much pets as i had.. heres the list

Slayers/Widow
220: 2590
217: 2561
213: 2324
209: 1957
205: 1750
200: 630

Dogs
219: 2834
215: 2348
211: 2295
207: 1915
203: 1729
175: 1270

Rihwen: 2240
Carlo: 1556-2300ish (more testing needed as apparently carlo seems to have an army of different weapons)

if sum1 could provide me with damage samples of crits on leet with and without pet snare(+100 AAO) i could calc the other ones too or u can do it urself using this:
BuffedPetCrit-BasePetCrit=DmgDifference
DmgDifference*4=CritMod
BasePetCrit/Critmod=Multipliers
Multipliers-1=ARMultipliers
ARMultipliers*400=PetAR
or use the formula
((BasePetCrit/((BuffedPetCrit-BasePetCrit)*4))-1)*400=PetAR


EDIT: updated list with Foreks calcs

[Forek]

ok so i decided to recalculate the AR of as much pets as i had.. heres the list

Slayers/Widow
220: 2590
217: 2561
213: 2324
209: 1957
200: 630

Dogs
219: 3091
215: 2348
211: 2295
207: 1915

I noticed that the AR values you calced were all slightly different to the numbers in the Engie Pet Attack Ratings thread. Do you still agree with the weapons that have previously been used to calculate those, or is it just rounding errors etc?

I've got all the lower slayers too so if no one else posts them before I can get on, I'll test them later this evening

[muha]

I prefer Mr_Stabby's calculation, from two reasons.

First, Occam's Razor. To the uninitiated, Occam's Razor is a scientific and philosophic rule that requires that the simplest of competing theories be preferred to the more complex or that explanations of unknown phenomena be sought first in terms of known quantities (Merriam-Webster Dictionary)

Now we know that a buffed widow misses a 3Kish evades player sometimes. That's quite easy to test. It was suggested that pet AR is high but maybe there is a special nerf here, or some other unknown. But here is a much simpler theory - AR was miscalculated. I prefer the simpler explanation.

Second, I think this makes less assumptions. The only assumptions I could find are:

• 400 AR = 1 multiplier. The evidence to this is abundant pre 1K AR.
• Pets have a single template, i.e. damage is linear with AR. I'm not sure what the evidence for this is, but I see no reason to assume otherwise.

In comparison, the old calculations also had to assume we know the weapon used. In Mr_Stabby's calculations the weapon type is only used as supportive evidence, and not as part of the argument.

I'm not sure what errors you refer to Rusty, can you be more specific? This seems pretty straightforward, unless a full day of teaching students made my brain melt (which is not such an impossibility).

[Forek]

Ok, apparently I don't have the 201 slayer, but I can add info for the 205 slayer and 175/203 dogs:

205 slayer

2973 - Unbuffed
3587 - Omni-Pol (+446 AR)
3911 - Omni-Pol + Sotos (+666 AR, +20 damage)

220 AR = +304
Base Damage = 553

=> 1750 AR


203 dog

2454 - Unbuffed
2968 - Omni-Pol (+446 AR)
3241 - Omni-Pol + Sotos (+666 AR, +20 damage)

666 AR = +767
Base Damage = 461

=> AR = 1729


175 dog

1712 - Unbuffed
2167 - Omni-Pol (+446 AR)
2414 - Omni-Pol + Sotos (+666 AR, +20 damage)

666 AR = +682
Base Damage = 410

=> AR = 1270

[Ebag333]

if sum1 could provide me with damage samples of crits on leet with and without pet snare(+100 AAO) i could calc the other ones too or u can do it urself using this:
BuffedPetCrit-BasePetCrit=DmgDifference
DmgDifference*4=CritMod
BasePetCrit/Critmod=Multipliers
Multipliers-1=ARMultipliers
ARMultipliers*400=PetAR
or use the formula
((BasePetCrit/((BuffedPetCrit-BasePetCrit)*4))-1)*400=PetAR

A very interesting thread!

I don't think this formula is quite right however. The numbers don't work for Rihwen when trying to use it to calculate MP's pet's damage.

Can you explain how you came up with this formula Stabby?

[PingBong]

just replace the number 4 in formula with whatever 400/AAOUbuffedPetWith equals aka if u buff pet with best anima only replace 4 with 1.32 and any1 done with highschool should be able to tell how this formula works

[Ebag333]

just replace the number 4 in formula with whatever 400/AAOUbuffedPetWith equals aka if u buff pet with best anima only replace 4 with 1.32 and any1 done with highschool should be able to tell how this formula works

Oh I'm done with highscool.

Dang, knew I shouldn't have taken Calc. Always knew that would drive out any knowledge of basic math out of my head.

Anyway, maybe it's a lack of punctuation, but I'm still not following how you got the formula.

Let me break it down step by step.

BuffedPetCrit-BasePetCrit=DmgDifference

Okay, this makes sense to me.

DmgDifference*4=CritMod

How do you know this is the CritMod?

BasePetCrit/Critmod=Multipliers

Multipliers-1=ARMultipliers

Where did "multipliers" come from (I assume you mean the result of the previous equation?), and why are we subtracting 1 from it?

ARMultipliers*400=PetAR

Where did 400 come from?


Additionally your formula doesn't match the breakdown above. You have:

((BasePetCrit/((BuffedPetCrit-BasePetCrit)*4))-1)*400=PetAR

When it should be (according to your breakdown):

(((BuffedPetCrit-BasePetCrit)*4)-1)*400=PetAR

So where did the BasePetCrit/ part come from?

Why is the order of operations different?


Not trying to be nitpicky here, but when doing calcs like this a single off number can skew the whole thing.

Additionally, wouldn't it be easier to work with min damage since they we don't have to take the crit modifier into effect? In my days of doing Calc (hey, my teacher WAS right, I will use this someday! ) we learned to always use a formula with the least number of variables where possible. If we use base damage then we have 3 variables, Base Damage, Pet Base AR, Buffed AR. If we use the crit multiplier we have actually 5. Base Damage, Pet Base AR, Buffed AR, Buffed Damage. Because crits a multiplier you can't simply subtract Buffed Damage to remove it, at least for MP pets because their AR buffs are tied with damage buffs. For engie pets I suppose you can because you're just using an AR buff, but you still have 4 variables as opposed to 3.

We also run into the additional problem that pets may be using a template of sorts.

I ran some tests of my own, and found that about 3.25 AR = 1 damage (it fluctuated by about +-.15, but I assume thats due to rounding).

Now whats interesting to me is that it doesn't change much. Allowing for rounding it very well could be 3.25 AR for every point of damage scaling to infinity.

But that's the intriguing part to me. It doesn't change.

This implies to me one of two scenario's is likely.

1) Pets don't have an AR template. No matter how their AR goes, they'll always get around 3.25 AR per point of damage (on MP pets at least).

2) MP pets aren't breaking the AR cap. Now since we don't really *know* our pets base AR, who's to say that +500 AR breaks whatever that cap is (seems likely it'd be around 2k, but that's a pure guess).

I really need to do some more tests, but I need to find a way of buffing my pet by more than just 500 AR. Unfortunatly only crats can help there, and their AR buffs aren't all that big. So if MP pets do have an AR cap, we may never be able to find it.

However, Engie pets should have a higher base AR (considering an unbuffed Chickenbot hits more often than a buffed Whiner in PVP ). Plus you get more AR buffs of different amounts, so if you stack them correctly you can get a very good sample size.

So you guys very well may run into an AR cap, which would go a long way to explaining why Chickenbot does less damage than Whiner/Carlos (since Rihwen is triple wielded and carlos is double or triple).

What I'd like to see is someone post some numbers from chicken bot with as many different combinations of AR buffs as possible, with min damage and min crit (210 Ado hecklers is what I used because they all have enough AC to reduce to min and are easily found).

Something like:

Unbuffed:
Min: XXXX
Min Crit: XXXX

+15 AR (Intrusive Aura of Entanglement):
Min: XXXX
Min Crit: XXXX

+30 AR (Intrusive Aura of Binding):
Min: XXXX
Min Crit: XXXX

+32 AR (Assist Combat Array):
Min: XXXX
Min Crit: XXXX

<skip a few>

+47 AR (Intrusive Aura of Entanglement, Assist Combat Array):
Min: XXXX
Min Crit: XXXX

Basically do combinations of those two buffs to get as small of increments as possible. With those numbers then we should be able to graph an AR curve (if there is one) and see if it changes over the +546 you can get. Oh, and if you *REALLY* want to get accurate, nab a crat and have them run THEIR AR buffs (of varying levels) to provide even more datapoints.

Your Intrusive Aura's (pet snares) would make excellent buffs, along with your damage boost buffs (IE: Omni pol) because neither buffs damage directly.

SotOS buffs damage directly, so makes it more difficult to determine AR.

Anyone feeling up to whacking on some hecks and recasting pet about 500 times?

[PingBong]

Oh I'm done with highscool.

i wouldnt count on that.. u just make no sense at all

for ur pleasure i had a 220 mp kick 2 leets, once unbuffed and second time with abomination and got hits 3293 unbuffed and 3940 buffed little calculation and we get 379 damage difference (with +dmg removed) from 304 AAO which makes 1 multiplier (400 aao) 499 damage. Divide base crit by 1 multiplier worth of critical hit and we find out rihwen has 6.6 multipliers, remove one due base weapon and get 5.6, multiply by 400 and get 2240 AR from unbuffed rihwen

[Ebag333]

That just doesn't make any sense at all.

To know your AR you have to know what weapon your using. Because AR adds to the base damage of the weapon.

Even IF your calculations are correct (which I'm not sure they are) without knowing the actual weapon used your AR calcs will be all wrong. You say "remove one due base weapon" I assume because you believe that the base weapon does 1 point of damage at QL1? Well the weapon that you believe the Widowmaker is using does 6 base damage at QL1. So you're thrown off right there.

Not to mention if you use damage buffs that damage you just added is multiplied which means you can't just simply subtract the damage. So your calcs will again be thrown off right there, unless you already know what the multiplier is. Which you don't.

Finally, why multiply by 400? Where do these numbers come from?

None of your numbers make a bit of sense. Sure they look pretty and you're throwing out a lot of numbers that seem to make sense, but when you get right down to it they just don't add up.

If I make the assumption that my pet is using a weapon with a base hit of 2240, then suddenly a 3k crit isn't that good. Not to mention he wouldn't have 2240 AR, he'd have more like 1 AR.

You can't just blatently make assumptions and base numbers off of that. If you want to seriously figure out the AR for the widowmaker, do the test that I suggested above. That will give you datapoints on the weapons actual base damage, as well as datapoints on the AR curve. From there you can calculate the AR curve in reverse to determine where base damage lies.

From there you'll need to parse the DB to find a weapon that matches the base damage (less the damage added from th AR of the pet).

Once you know the actual weapon, then you can determine what the pets actual base AR is.

Once you know the base AR, then you're set for calculating all sorts of things, like exactly how much one buff will give the pet.

If I had a 220 engie I'd get those datapoints myself, but I don't. As I mentioned before because the MP AR buffs add damage it makes it much more difficult to plot an AR curve (additionally we simply have less AR buffs so would have very few datapoints in comparison, which could skew any sort of AR curve we try and generate).

Without doing that you'll never know if the pet has an AR template, which also could be throwing off your calculations, and thus any number you pick is just a random number out of thin air.

So. Any volunteers? If you're on RK2 I'd be happy to lend what assistance I can, and if you're not on RK2 I'll be happy to help chart and calculate the AR curves based on the datapoints.

/me edits

I was just re-reading through this and realized I'd completely missed something.

ok so i decided to recalculate the AR of as much pets as i had.. heres the list

Slayers/Widow
220: 2590
217: 2561
213: 2324
209: 1957
205: 1750
200: 630

Are you seriously suggesting that the level 200 pet has 630 AR?

First of all, the 200 pet doesn't use the same weapons as the SL pets (and furthermore the SL pets may use different weapons than the chickenbot...in fact, for all we know each pet uses a completely different weapon, or there might be just one pet that's bugged and has the wrong weapon).

Secondly, even if it did have 630 AR (which it doesn't) with 630 AR it'd never land hits on Ado+ hecks. Which it doesn't. In fact, I seem to recall that people use the 200 Slayer in Ado+ for a reason, which is certainly not because it misses all the time.

[theobelisk]

i dont even have an engi and i know the 200 pet od's the sl pets up until +209 or so. and the 631 ar is unfbuffed, it hits like 1.1k buffed if i recall, which is enough

[Ebag333]

Sorry for the double post, but the last one was getting long and I wanted to address Muha's post.

Now we know that a buffed widow misses a 3Kish evades player sometimes. That's quite easy to test. It was suggested that pet AR is high but maybe there is a special nerf here, or some other unknown. But here is a much simpler theory - AR was miscalculated. I prefer the simpler explanation.

There are a few problems with this assumption.

We don't really know exactly how AAD is calculated VS evades. A pet will never miss a mob (or at least I've never seen mine miss), yet they miss players with around 2k evades on a regular basis (at least MP pets do).

AAD must not be 100% equal to evadesm because otherwise the pets should be missing mobs, which is something that I do with 2k AR on a regular basis (as well as missing players! ).

Personally, I believe that for most purposes AAO = AAD, and Weapon/Nano Skill = Evades/NR. This is the simplest explanation (Occam's Razor anyone?) and goes a long way to explaining why certain people can land certain things on low evade high AAD toons (IE: crats) but can't land on high evade low AAD toons.

This is further evidenced by the fact that my pet (even with the newest buffs) will struggle to land anything on a MA or Advy (a high evade relatively low AAD toon) but will land hits fairly easily on a crat (relatively low evade high AAD toon).

I speculate (and this part here is just pure speculation) that the pets must have relatively low weapon skill (perhaps 0, which would make sense considering that trader debuffs had no effect on mobs until they added in the -AAO) and a high AAO skill. If this is in fact true, then it goes a long way to explaining a lot of different things.

Of course we'll probably never know how Weapon/Nano skill, AAO, AAD, and evades are calculated. But there's a lot we can glean from looking at the numbers.

Second, I think this makes less assumptions. The only assumptions I could find are:

• 400 AR = 1 multiplier. The evidence to this is abundant pre 1K AR.
• Pets have a single template, i.e. damage is linear with AR. I'm not sure what the evidence for this is, but I see no reason to assume otherwise.

The problem with these assumptions is:
1) the first one is based off player AR. I've found in my own tests that for pets it's quite different.

2) We have no evidence for or against pets using a single template. For all we know pets AR has a template up to a certian point, then just stops (think of MA's a while back). Or it could have a cut off point where it changes drastically (IE: MP's, traders, etc). There's no reason to assume that they don't have a single template, but there's no reason to assume they do either.

In comparison, the old calculations also had to assume we know the weapon used. In Mr_Stabby's calculations the weapon type is only used as supportive evidence, and not as part of the argument.

I'm not sure what errors you refer to Rusty, can you be more specific? This seems pretty straightforward, unless a full day of teaching students made my brain melt (which is not such an impossibility).

Full day of teaching would make anyone's brain melt.

You still have to know the weapon in order to be able to determine the numbers correctly.

With the method I suggested above you could find a weapon that *MIGHT* match the damage/AR curve at one given point. Of course that's only one point, you'd have to calculate that weapon with the appropriate amounts (determined by the AR curve you already know) and see if that weapons AR curve follows what you have.

If it doesn't, then it's not the right one, if it is, then you've found it.

But without someone getting those datapoints (and the more the better because we really don't know where that base AR is, so we have no idea if we're hitting a cap somewhere or if it even is a straight line) there's absolutely no way of telling what weapon is being used, short of pulling a random one out of the DB, and checking to see if the expect damage curve matches that of what you get.

Needless to say with the thousands of monster weapons and whatnot in DB, that's essentially an impossible task.

Long story short, gotta do your homework first before you know the answers to the questions on the test.

[Forek]

Ebag333, what are you smoking?

[Ebag333]

i dont even have an engi and i know the 200 pet od's the sl pets up until +209 or so. and the 631 ar is unfbuffed, it hits like 1.1k buffed if i recall, which is enough

I'm not saying that it has 631 AR. I'm saying two things.

1) You can't use the same calculations because the RK slayer and SL pets use different weapons, with different stats. Oh sure, if you knew the actual formula used by FC to calculate it, then you could swap it around until you could plug in a few numbers and figure it out. But since we're essentially reverse engineering it, the formula's we use to calculate AR, damage, etc, that formula is only good for that one specific datapoint.

2) With 631 AR the Slayer wouldn't hit squat anywhere. Unless of course you're suggesting that 250+ mobs in inferno only have around 600 AAD? (Since mobs don't have evades.)

[Mailana]

I'm not saying that it has 631 AR. I'm saying two things.

1) You can't use the same calculations because the RK slayer and SL pets use different weapons, with different stats. Oh sure, if you knew the actual formula used by FC to calculate it, then you could swap it around until you could plug in a few numbers and figure it out. But since we're essentially reverse engineering it, the formula's we use to calculate AR, damage, etc, that formula is only good for that one specific datapoint.

2) With 631 AR the Slayer wouldn't hit squat anywhere. Unless of course you're suggesting that 250+ mobs in inferno only have around 600 AAD? (Since mobs don't have evades.)

The RK slayer gets a trimmer and 3 buffs which add up to around 800 ar based on the engies level. Thats why people say 4k ar even tho our highest pets spawn with closer to 3k. Your assumption is similar to saying that docs should never even try to wield a weapon, since without trickle imps buffs and gear they can only get 600 skill.

As to his calculations, I havent done the math but the concept seems to be an improvement over the old ones, which made assumptions based on one datapoint and applied it to all the others (the weapon item in database). You CAN use the same calculations, since the weapon is a variable but the ar curve is not. +400 ar = x1 multiplier is a basic rule that applied to everyone exactly pre-sl, and seems to still apply to pets and mobs.

[PingBong]

for Ebag666:
U dont need to know the weapon the pets use, all u need is to find out how the AR affects the damage as damage increase due to ar is always the same on pets no matter how high it is, multipliers are just the catalyst that translate AR increase into damage and 400 ar=1 multiplier, why i remove 1 multiplier each time? because when wielding a weapon with 0 ar you still have 1 multiplier applied to the weapon by default, if u have 400 ar u have 2 multipliers. To prove that pets still use 400ar to 1 multiplier conversion i will now predict widowmaker critical hits with different buffs running on it:
what we have - 2590 base AR as a result of my previous calculations and 689 damage increase over 400 AR
Omni pol: (((2590+446)/400)+1)*689=5919
Widowmaker Battle Drone hit 34-I Helper for 5919 points of melee damage.Critical hit!
omni pol+snare: (((2590+446+100)/400)+1)*689=6091
Widowmaker Battle Drone hit Biofreak Carcass for 6091 points of melee damage.Critical hit!
omni pol+snare+sotos: (((2590+446+100+220)/400)+1)*689+20=6490
Widowmaker Battle Drone hit 34-I Helper for 6490 points of melee damage.Critical hit!
consistent no? if u need anything more toss me a list of different evocs for example so i can predict the damage done by rihwen with them running and u can check it out urself..

[Ferg]

Yeah, my testing is showing the same results so I guess the chicken isn't as uber as everyone thinks it is. This means the chicken only has a max of 29 AR more than the 217, thats pathetic. My calculation actually brought the 217 in at 2572 AR (probably because i didn't drop the decimal) which would only be a difference of 18. I can't wait til the crats and mps hear this one :P

[Technogen]

What weapon does the dog have? remember it's got that templet problem if they come out with the same weapon that would be hard proof that it has more ar than the widow.

oh and Ebag RK slayer misses a good deal, why it doesnt od all the sl pets out right.

[LordArthur]

This makes sense to me.. Lately the pets have seemed to hit less on most profs, and with more evades/aad coming out everyday I can only assume it will get worse...

We need another "any engi that took the time and effort to lvl to 220 deserves better"-post from devs

[Rustybolts]

I'm convinced. Good work Stabby